Cyclometallated iridium complexes have been the focus of research and development in OLED (Mark E. Thompson et al, WO 01/41512 A1) display devices over last several years. Those complexes can offer higher efficiency when used as phosphorescent dopants in OLED devices since both singlet and triplet excitons generated by electroexcitation can be harvested by a phosphorescent dopant, while only singlets (25% of total excitons) can be utilized when a fluorescent material is used as a dopant. Tris-cyclometallated iridium complexes have demonstrated such advantage. There exist two stereoisomers in homoleptic tris-cyclometallated iridium complexes such as tris(2-(phenyl)pyridinato-N,C2)iridium (III) (Ir(ppy)3), namely facial and meridional isomers as shown below. The facial isomer has been shown to be more desirable as it has demonstrated higher quantum yield and thermal stability than the corresponding meridional isomer (A. B. Tamayo, et al, J. Am. Chem. Soc. 2003, 125, 7377).

There are continuous efforts to develop new phosphorescent dopants for improving the efficiency and operational stability of OLED devices. Heteroleptic (mixed) tris-cyclometallated iridium complexes have recently attracted attention of research community and their applications to OLED devices have been demonstrated (T. Igarashi et al, US 2001/0019782 A1; J. Kamatani, et al, US 2003/0068526 A1; S. Akiyama et al, JP2003-192691A). However, the synthesis of those heteroleptic complexes is challenging. The method employed in the prior arts involves the reaction of a bis-cyclometallated iridium complex with a third ligand in glycerol at high temperature (usually above 180° C.), which we found to produce a mixture of different homoleptic and heteroleptic tris-cyclometallated iridium complexes formed from ligand-scrambling side reactions and lead to difficulties in separation and purification of desired compounds. Recently, we have developed a novel method to prepare mixed tris-cyclometallated iridium complexes in high yields and purity, but the products obtained from this reaction are meridional isomers (S. Huo, U.S. Pat. No. 6,835,835). We also discovered that some meridional isomers could isomerize to their facial isomers by applying heat to a solution of the meridional isomer in DMSO, but it was accompanied by severe decomposition in some cases. Another method for this isomerization involves the use of an acid and silica gel particles US2006/0135772. Although the method allowed isolation of pure facial isomer readily, the yield of the product was not satisfactory. Thompson et al reported thermal (by refluxing the meridional isomer in glycerol) or photochemical isomerization of homoleptic meridional tris-cyclometallated iridium complexes (Tamayo et al, J. Am. Chem. Soc. 2003, 125, 7377-7387). A photochemical isomerization of homoleptic tris-cyclometallated iridium complexes from meridional isomers to facial isomers is also disclosed in a patent application (JP 2004189673 A2). However, we found that the methods did not work for some heteroleptic tris-cyclometallated iridium complexes such as meridional bis-(1-(phenyl)isoquinolinato-N, C2)(phenylpyridinato-N,C2)iridium (mer-Ir(1-piq)2(ppy)). For example, thermal isomerization of mer-Ir(1-piq)2(ppy) in glycerol under the same conditions described in literature (Tamayo et al, J. Am. Chem. Soc., 2003, 125, 7377-7387) resulted in largely decompositions and severe ligand-scrambling. Photo irradiation of mer-Ir(1-piq)2(ppy) did not produce the corresponding facial isomer. Moreover, the photochemical process may not be suitable for large scale productions and the use of glycerol is not convenient for isolation and purification of the product.
It is a problem to be solved to provide a simple and efficient process for the isomerization of meridional tris-cyclometallated iridium complexes to their facial isomers.